Extreme ultraviolet lithography (EUV, also known as soft x-ray projection lithography) is a contender to replace deep ultraviolet lithography for the manufacture of 0.13 micron, and smaller, minimum feature size semiconductor devices.
However, extreme ultraviolet light, which is generally in the 5 to 40 nanometer wavelength range, is strongly absorbed in virtually all materials. For that reason, extreme ultraviolet systems work by reflection rather than by transmission of light. Through the use of a series of mirrors, or lens elements, and a reflective element, or mask blank, coated with a non-reflective absorber mask pattern, the patterned actinic light is reflected onto a resist-coated semiconductor wafer.
The lens elements and mask blanks of extreme ultraviolet lithography systems are coated with reflective multilayer coatings of materials such as molybdenum and silicon. Reflection values of approximately 65% per lens element, or mask blank, have been obtained by using substrates that are coated with multilayer coatings that strongly reflect light essentially at a single wavelength within an extremely narrow ultraviolet bandpass; e.g., 12 to 14 nanometer bandpass for 13 nanometer ultraviolet light.
There are various classes of defects in semiconductor processing technology which cause problems. Opaque defects are typically caused by particles on top of the multilayer coatings or mask pattern which absorb light when it should be reflected. Clear defects are typically caused by pinholes in the mask pattern on top of the multilayer coatings through which light is reflected when it should be absorbed. And phase defects are typically caused by scratches and surface variations beneath the multilayer coatings which cause transitions in the phase of the reflected light. These phase transitions result in light wave interference effects which distort or alter the pattern that is to be exposed in the resist on the surface of the semiconductor wafer. Because of the shorter wavelengths of radiation which must be used for sub-0.13 micron minimum feature size, scratches and surface variations which were insignificant before now become intolerable.
While progress has been made in reducing or eliminating particle defects and work has been done on repair of opaque and clear defects in masks, to date nothing has been done to address the problem of phase defects. For deep ultraviolet lithography, surfaces are processed to maintain phase transitions below 60 degrees. Similar processing for extreme ultraviolet lithography is yet to be developed.
For an actinic wavelength of 13 nanometers, a 180 degree phase transition in the light reflected from the multilayer coating may occur for a scratch of as little as 3 nanometers in depth in the underlying surface. This depth gets shallower with shorter wavelengths. Similarly, at the same wavelength, surface variations more abrupt than one (1) nanometer rise over one hundred (100) nanometers run may cause similar phase transitions. These phase transitions can cause a phase defect at the surface of the semiconductor wafer and irreparably damage the semiconductor devices.
In the past, mask blanks for deep ultraviolet lithography have generally been of glass but silicon or ultra low thermal expansion materials have been proposed as alternatives for extreme ultraviolet lithography. Whether the blank is of glass, silicon, or ultra low thermal expansion material, the surface of the mask blank is made as smooth as possible by such processes a chemical mechanical polishing, magneto-rheological finishing, or ion beam polishing. The scratches that are left behind in such a process are sometimes referred to as “scratch-dig” marks, and their depth and width depend upon the size of the particles in the abrasive used to polish the mask blank. For visible and deep ultraviolet lithography, these scratches are too small to cause phase defects in the pattern on the semiconductor wafer. However, for extreme ultraviolet lithography, scratch-dig marks are a significant problem because they will appear as phase defects.
Due to the short illumination wavelengths required for EUV lithography the pattern masks used must be reflective mask instead of the transmissive masks used in current lithography. The reflective mask is made up of a precise stack of alternating thin layers of molybdenum and silicon, which creates a Bragg refractor or mirror. Because of the nature of the multilayer stack and the small feature size, any imperfections in the surface of the substrate on which the multilayer stack is deposited will be magnified and impact the final product. Imperfections on the scale of a few nanometers can show up as printable defects on the finished mask and need to be eliminated from the surface of the mask blank before deposition of the multilayer stack.
Common imperfections include pits, scratches, and particles. Common cleaning techniques remove many of the particles but either generate new pits or amplify existing pits. The pits can come from the polishing or cleaning process or can be from inclusions or flaws in the substrate material itself that are exposed during the cutting and polishing process. Further polishing can be used to remove the pits at the surface, but there is a risk that new pits will be exposed or caused in the process, which limits the usefulness of using polishing alone to smooth and planarize the substrate surface. Another method for substrate smoothing is laser or plasma annealing. These techniques melt and reflow a thin surface layer of the glass substrate, removing local defects. The problem is that they induce longer range roughness or ripples in the substrate surface and so do not provide the substrate flatness required for EUV mask blanks.
In view of the need for the increasingly smaller feature size of electronic components, it is increasingly critical that answers be found to these problems. In view of the ever-increasing commercial competitive pressures, along with growing consumer expectations, it is critical that answers be found for these problems. Additionally, the need to reduce costs, improve efficiencies and performance, and meet competitive pressures adds an even greater urgency to the critical necessity for finding answers to these problems.
Solutions to these problems have been long sought but prior developments have not taught or suggested any solutions and, thus, solutions to these problems have long eluded those skilled in the art.